eYARD: An introduction to AI

As many of you might know, a couple of weeks ago we ran our first ever webinar; ‘What is AI and what can it do for your terminal?’. This was an exciting opportunity to introduce eYARD to those who joined us in the session. However, as we appreciate that not everyone would have been able to join on the day, we did not want you to miss out on hearing about how this project can support you and so if this is something that intrigues you, read on –

In conversation with Pablo Fernandez-Peña and Mike Dekker, we looked back at the way port terminal operations were handled in recent years and the positive impact the introduction of AI could have in this sector. We heard of cases in which AI is implemented to predict earthquakes or protect wildlife and deforestation. If we allowed machinery to also study our terminal processes; wouldn’t we have the option of saving time from unproductive moves and reducing cost in doing so? And with that in mind, we learned AI is here to stay.

93% of our attendees confirmed that – in their experience – terminal unproductive moves result in additional costs and so we thought it was important to explore how AI would be applied. eYARD would study two components; existing data and predictability regardless of the nature of the terminal. By using reinforcement learning algorithms, this technology considers all stacking possibilities based on the existing terminal layout and redefines transit to achieve KPI efficient results and optimal operational management.

At a proof of concept stage and in using emulation, eYARD replays the terminal history with existing quality data over a specific period. At this stage, the algorithm is not being used but through emulation it is defining the zero-estate value of its KPIs. Afterwards, the algorithm is used to determine where containers are going to be positioned during the operation. These last results are compared to the original values and adjusted – in the event of any parameters needing to be changed to adjust these to the most desirable outcome.

As long as the above data can be provided and the existing TOS allows for communications with third party applications, eYARD can be integrated with any container yard or equipment. By tapping into the existing TOS via an API protocol and as a Cloud-base solution, this technology maintains the simplicity and promotes usability of such a tool. After the integration has taken place, the operating team will have visibility of a dashboard that can be customized to suit the needs of its users. That is, the data that senior management requires will potentially vary from the data that another technical department might find useful (i.e. IT) within the same organisation.

Over upcoming webinars we will continue to share snippets of what this data science revolution has to offer in this particular sector. We look forward to having you join us in the future.

FEB. 10, 2021